Google Real-world Vulnerability Discovery
Progress Over Time
Interactive timeline showing model performance evolution on Google Real-world Vulnerability Discovery
Google Real-world Vulnerability Discovery Leaderboard
| Context | Cost | License | ||||
|---|---|---|---|---|---|---|
| 1 |  Google | — | — | — |
What is Google Real-world Vulnerability Discovery?
Google's internal vulnerability-discovery benchmark measures recall across more than 1,200 recent confirmed vulnerabilities in popular open-source projects spanning 20 programming languages.
Google Real-world Vulnerability Discovery is a text benchmark evaluating models on safety, agents, and code tasks. LLM Stats tracks 1 models on this benchmark, scored on a 0–1 scale. The current average is 0.7, with the leader at 0.7.
Compare leaders on the best AI for safety, best AI for agents and best AI for code leaderboards.
Current leaders
Gemini 3.8 Flash Cyber from Google currently leads the Google Real-world Vulnerability Discovery leaderboard with a score of 0.710 across 1 evaluated AI models.
FAQ
Common questions about the Google Real-world Vulnerability Discovery benchmark and leaderboard.